Generative modeling becomes increasingly data-intensive in high-dimensional spaces. In molecular science, where data collection is expensive and important events are rare, compression to lower-dimensional manifolds is especially important for various downstream tasks, including generation. We combine a time-lagged information bottleneck designed to characterize molecular important representations and a diffusion model in one joint training objective. The resulting protocol, which we term Diffusive State Predictive Information Bottleneck (D-SPIB), enables the balancing of representation learning and generation aims in one flexible architecture. Additionally, the model is capable of combining temperature information from different molecular simulation trajectories to learn a coherent and useful internal representation of thermodynamics. We benchmark D-SPIB on multiple molecular tasks and showcase its potential for exploring physical conditions outside the training set.
@article{arxiv.2510.09784,
title = {Combined Representation and Generation with Diffusive State Predictive Information Bottleneck},
author = {Richard John and Yunrui Qiu and Lukas Herron and Pratyush Tiwary},
journal= {arXiv preprint arXiv:2510.09784},
year = {2025}
}